A 30% strike rate sounds good. A 40% strike rate sounds better. But a punter who strikes at 40% on favourites averaging $1.90 is going broke, while a punter who strikes at 15% on roughies averaging $9.00 is building a bankroll.
The difference is expected value — EV. It's the mathematical lens that separates gambling from investing, noise from edge, and profitable punters from the crowd. This article defines EV, shows you how to calculate it, and explains why strike rate is a trailing indicator that tells you nothing about whether your next bet is +EV.
If you absorb one concept from this entire series, make it this one.
What Is Expected Value?
Expected value is the average amount you expect to win or lose per dollar staked if you could repeat a bet an infinite number of times under identical conditions.
EV = (Probability of Win × Profit per Win) − (Probability of Loss × Stake)
Where:
- Probability of win — your estimate of the horse's true chance (as a decimal).
- Profit per win — the net return if the bet wins (odds minus 1, times stake).
- Probability of loss — 1 − probability of win.
- Stake — your bet size.
A positive EV (+EV) means you expect to profit long-term. A negative EV (−EV) means you expect to lose. Zero EV means break-even in expectation.
EV is not about what happens on the next bet. It's about what happens over the next thousand.
A Worked Example: Calculating EV on a Single Bet
You're looking at a horse at Randwick priced at $5.00. After analysing the race, you estimate it has a 22% chance of winning. You plan to stake $100.
Profit per win = ($5.00 − 1) × $100 = $400
Probability win = 0.22 Probability loss = 0.78
EV = (0.22 × $400) − (0.78 × $100)
= $88 − $78
= +$10
In expectation, this bet returns $10 for every $100 staked — a 10% edge. Over 100 similar bets you'd expect to be up roughly $1,000 in gross terms, though variance means the actual result rarely matches the expectation exactly.
Now the same horse at $3.50, with your estimate unchanged at 22%:
Profit per win = ($3.50 − 1) × $100 = $250
EV = (0.22 × $250) − (0.78 × $100)
= $55 − $78
= −$23
Negative. You expect to lose $23 per $100. Your read of the horse hasn't changed — only the price has. The price is everything.
The EV Formula in Percentage Terms
Professionals usually express EV as a percentage of stake — it makes bets of different sizes comparable and edge easy to track:
EV% = (Probability of Win × Decimal Odds) − 1
Using the two examples above:
(0.22 × $5.00) − 1 = +10%
(0.22 × $3.50) − 1 = −23%
A +10% EV bet returns 110% of stake in the long run; a −23% EV bet returns 77%. EV% is the metric EVSTREAM displays beside every runner on the grid — when a cell highlights green, it's flagging a price where the fair probability implies an edge over the bookmaker's published odds.
Why Strike Rate Is Misleading
Strike rate is the percentage of your bets that win. It's a historical fact, not a predictive tool — and it's dangerously seductive.
A punter striking at 35% feels skilled: more than one in three bets win. But if their average winning price is $2.40, their return is 35% × $2.40 = $0.84 per dollar — losing 16% on turnover. High strike rate, negative EV.
Flip it: a punter striking at 12% with an average winning price of $10.00 returns 12% × $10.00 = $1.20 per dollar — winning 20% on turnover. Low strike rate, positive EV.
| Punter | Strike rate | Avg winning odds | Return per $1 | EV status |
|---|---|---|---|---|
| A | 35% | $2.40 | $0.84 | −16% (losing) |
| B | 22% | $4.50 | $0.99 | −1% (near break-even) |
| C | 12% | $10.00 | $1.20 | +20% (profitable) |
Punter A has the highest strike rate and the worst results. Punter C has the lowest and the best. The market doesn't reward frequency. It rewards pricing.
Probability Is the Input, Not the Output
The confusion around strike rate comes from a backwards causal chain. Punters think: pick winners → winners produce profit.
The correct chain runs the other way:
Estimate probabilities accurately → combine with favourable odds → +EV → repeated over time → profit → and a strike rate falls out as a side effect.
Your probability estimate is the input. The odds you accept are the second input. EV is the calculation that combines them. The result of any single bet is variance; the long-term result converges on EV. Strike rate is the statistical residue of that process.
If you don't estimate probabilities before you bet, you aren't calculating EV. If you aren't calculating EV, you aren't betting — you're guessing.
The EV Spectrum
Every bet sits somewhere on the EV spectrum. Knowing where is the first step to moving rightward.
−30% −20% −10% 0% +5% +15%
│ │ │ │ │ │
MUG AVERAGE SHARP FAIR SOLID STRONG
BETS PUNTER RECREATIONAL MARKET EDGE EDGE
- Mug bets (−30% and worse). Same-race multis, most exotics, and bets into 140%+ overround markets where no edge exists. The domain of promo-chasers who don't understand margin.
- Average punter (−20% to −10%). Some form knowledge, reasonable horses, but never calculates margin or EV. Beats the random bettor, loses to the market.
- Sharp recreational (−5% to 0%). Understands form, shops for price, may break even against published odds — but still pays margin. The gap to profit is usually de-vigging and fair-odds discipline.
- Fair market (0%). Break-even. A coin flip at fair odds.
- Solid edge (+5% to +10%). Professional territory: consistent value, bankroll management, respect for variance. Where sustainable profit lives.
- Strong edge (+15%+). Rare — niche markets, slow-adjusting early markets, promo overlays. Hard to sustain at scale, highly profitable when available.
Most punters live on the left and don't know it. The move right requires one thing: calculating EV before every bet.
The Math Behind the Long Run
Why does EV matter if individual bets are random? The law of large numbers converges your actual results toward your expected results over time. Place 1,000 bets at an average +5% EV and your expected profit is 5% of turnover — the actual figure varies, but the distribution centres on +5%, and its spread shrinks relative to sample size.
This is why bankroll management matters so much. A +EV bettor with poor staking can be wiped out by variance before the long run arrives; a −EV bettor with good staking just loses more slowly. Only +EV combined with appropriate stakes produces sustainable profit. (The Kelly Criterion, covered later in the practical module, derives optimal stakes directly from EV and probability — no EV, no Kelly.)
EV in the Context of Fair Odds
EV is the gap between the odds offered and the fair odds implied by true probability. If fair odds for a horse are $5.00 (a 20% chance) and you're offered $6.00:
(0.20 × $6.00) − 1 = +20%
Offered $4.00 instead:
(0.20 × $4.00) − 1 = −20%
The fair odds are your anchor; the published odds are the price you pay; EV states whether that price is favourable. This is why fair odds are a prerequisite — you can't know a bet is +EV without the true probability, and you can't get the true probability without stripping margin from the market:
published odds → implied probability → de-vig → fair probability → compare to your estimate → calculate EV → bet only if +EV
Skip any step and the chain breaks.
Why Professionals Calculate EV Before Every Bet
Professionals don't bet because they "like" a horse. They bet because the EV calculation justifies the stake — the horse is a vehicle for a +EV position. If the price moves against them, the bet is no longer +EV and they pass, even if their form read still favours the horse.
The enthusiast backs their opinion. The professional backs the price. The opinion informs the probability estimate; the price determines the action.
When EVSTREAM highlights a runner green, it's running exactly this calculation in real time: estimating fair probability from multiple market sources, comparing it to the published odds at 60+ bookmakers, and surfacing +EV opportunities as they appear. The 0.4-second tick matters because edges are fleeting — the price you saw ten seconds ago may not be +EV now.
Key Takeaways
- EV% = (win probability × decimal odds) − 1. It measures expected return per dollar staked.
- +EV bets profit in expectation; −EV bets lose. Outcome on any single bet is variance.
- Strike rate is an output, not an input — set by your odds and edge, not willpower.
- High strike rate at short odds can lose; low strike rate at long odds can win.
- Fair odds are required to calculate EV — without true probability you can't know if a bet is +EV.
- Professionals bet the price, not the horse.
See Live EV on Every Runner
You don't have to run this by hand. EVSTREAM shows fair odds, published odds and EV% beside every runner, updated every 0.4 seconds across every bookmaker and code — and highlights +EV cells the moment a price creates an edge.
Start your 7-day free trial to see live EV on the grid, or open the grid and watch the green cells appear as the market moves.
